sacred

Sacred 安装

# 主角
pip install sacred

# 用于数据库连接
pip install numpy pymongo

MongoDB 安装

MongoDB 是一个数据库管理系统，这里用作 Sacred 的存储后端。

在 ubuntu 上的 MongoDB 安装可以参考 Install MongoDB Community Edition on Ubuntu，其他系统也可以在该网站上找到对应的安装方式。

mongoDB 常用命令

# 启动
sudo service mongod start

# 停止
sudo service mongod stop

# 重启
sudo service mongod restart

# 进入MongoDB
mongosh

创建一个名为 sacred 的数据库，用作 sacred 工具的后端存储：

# 进入MongoDB
mongosh

# 创建sacred数据库。use命令切换数据库，没有该数据就会自动创建一个
use sacred

Omniboard安装

在 Ubuntu 机器上安装版本≥v8 的 Node.js，系统默认 apt 安装的版本不够，需要手动安装

安装node.js和npm

sudo apt install nodejs
sudo apt install npm

新版本安装(hexo不支持)

curl -fsSL https://deb.nodesource.com/setup_20.x | sudo -E bash - &&\
sudo apt-get install -y nodejs

nvm指定版本安装

proxychains4 wget -qO- https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.3/install.sh | proxychains4 bash
source ~/.zshrc
export NVM_NODEJS_ORG_MIRROR=https://npm.taobao.org/mirrors/node/
export NVM_IOJS_ORG_MIRROR=http://npm.taobao.org/mirrors/iojs
nvm install 12.16.2
nvm use 12.16.2

换源

sudo npm install -g nrm
nrm ls
nrm use taobao

# 4. 测试安装版本信息
node -v
npm version
npx -v

第二步，npm 安装 omniboard

npm install -g omniboard

第三步，开启 omniboard 服务。平时也是用该命令开启 omniboard 可视化前端

# 开启用法
omniboard -m hostname:port:database

# 默认情况下如下，其中27017是MongoDB的端口
omniboard -m localhost:27017:sacred

第四步，打开 http://localhost:9000 来查看前端，并进行管理。

添加远程查看

ssh -L 9000:127.0.0.1:9000 -p 10102 [email protected]

使用案例

使用 yunjey 的一个 pytorch 教程作为演示，代码是演示用 pytorch 实现基于 CNN 的 MINIST 手写数字识别。

根据 Sacred 文档稍作修改，就可以演示如何进行实验的记录。

更多用法请去看 Sacred 文档：Welcome to Sacred’s documentation!。内容超丰富，功能超级多。

代码：

from sacred import Experiment
from sacred.observers import MongoObserver
from sacred.utils import apply_backspaces_and_linefeeds

import torch
import torch.nn as nn
import torchvision
import torchvision.transforms as transforms


ex = Experiment("mnist_cnn")
ex.observers.append(MongoObserver.create(url='localhost:27017',
                                         db_name='sacred'))
ex.captured_out_filter = apply_backspaces_and_linefeeds


# 超参数设置
@ex.config
def myconfig():
    # Device configuration
    device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')

    # Hyper parameters
    num_epochs = 5
    num_classes = 10
    batch_size = 100
    learning_rate = 0.001


# Convolutional neural network (two convolutional layers)
class ConvNet(nn.Module):
    def __init__(self, num_classes=10):
        super(ConvNet, self).__init__()
        self.layer1 = nn.Sequential(
            nn.Conv2d(1, 16, kernel_size=5, stride=1, padding=2),
            nn.BatchNorm2d(16),
            nn.ReLU(),
            nn.MaxPool2d(kernel_size=2, stride=2))
        self.layer2 = nn.Sequential(
            nn.Conv2d(16, 32, kernel_size=5, stride=1, padding=2),
            nn.BatchNorm2d(32),
            nn.ReLU(),
            nn.MaxPool2d(kernel_size=2, stride=2))
        self.fc = nn.Linear(7 * 7 * 32, num_classes)

    def forward(self, x):
        out = self.layer1(x)
        out = self.layer2(out)
        out = out.reshape(out.size(0), -1)
        out = self.fc(out)
        return out


# notice how we can access the message here by taking it as an argument
@ex.automain
def main(device,num_epochs,num_classes,batch_size,learning_rate ):
    # MNIST dataset
    train_dataset = torchvision.datasets.MNIST(root='/home/ubuntu/Datasets/MINIST/',
                                               train=True,
                                               transform=transforms.ToTensor(),
                                               download=True)

    test_dataset = torchvision.datasets.MNIST(root='/home/ubuntu/Datasets/MINIST/',
                                              train=False,
                                              transform=transforms.ToTensor())

    # Data loader
    train_loader = torch.utils.data.DataLoader(dataset=train_dataset,
                                               batch_size=batch_size,
                                               shuffle=True)

    test_loader = torch.utils.data.DataLoader(dataset=test_dataset,
                                              batch_size=batch_size,
                                              shuffle=False)

    model = ConvNet(num_classes).to(device)

    # Loss and optimizer
    criterion = nn.CrossEntropyLoss()
    optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate)

    # Train the model
    total_step = len(train_loader)
    for epoch in range(num_epochs):
        for i, (images, labels) in enumerate(train_loader):
            images = images.to(device)
            labels = labels.to(device)

            # Forward pass
            outputs = model(images)
            loss = criterion(outputs, labels)

            # Backward and optimize
            optimizer.zero_grad()
            loss.backward()
            optimizer.step()

            if (i + 1) % 100 == 0:
                print('Epoch [{}/{}], Step [{}/{}], Loss: {:.4f}'
                      .format(epoch + 1, num_epochs, i + 1, total_step, loss.item()))

    # Test the model
    model.eval()  # eval mode (batchnorm uses moving mean/variance instead of mini-batch mean/variance)
    with torch.no_grad():
        correct = 0
        total = 0
        for images, labels in test_loader:
            images = images.to(device)
            labels = labels.to(device)
            outputs = model(images)
            _, predicted = torch.max(outputs.data, 1)
            total += labels.size(0)
            correct += (predicted == labels).sum().item()

        print('Test Accuracy of the model on the 10000 test images: {} %'.format(100 * correct / total))

    # Save the model checkpoint
    torch.save(model.state_dict(), 'model.ckpt')

执行完该程序后，可以打开 omniboard 前端 http://localhost:9000